State Estimator Observability and Redundancy Requirements

State Estimator Observability and Redundancy Requirements

Proposal revised and endorsed by Reliability and Operations Subcommittee

October 12, 2004

1

State Estimator Observability and Redundancy Requirements

ROS endorsed 10/12/2004

Purpose

Executive Summary

1.  Introduction

1.1  Operating States of a Power System

1.2  Why State Estimator?

1.3  Why RTCA? (Real time Contingency Analysis)

1.4  Model of ERCOT System

1.5  Cost vs ERCOT Responsibility

2.  Observability Analysis – How and Results

3.  Redundancy and Critical Measurements

4.  Topology, Validity and Availability

5.  Observability Requirement & available exceptions

6. Redundancy Standard and Exceptions

Appendix 1

1.  State Estimator basics

1.1  Network Topology processing (NTP)

1.2  Measurement Consistency Checking

1.3  Observability Analysis

1.4  WLS State Estimator

1.5  Bad Data Detection and Filtering

2.  Topology Estimator

3.  Time Schedule Model Implementations

4.  Manual Interventions in the state estimator

Appendix 2

Observability and Optimal RTU Placement

Appendix 3

RTU Placement Results

Purpose

Over the last year ERCOT has implemented the use of Real Time Contingency Analysis (RTCA) as the primary real-time reliability tool for transmission security. This tool is highly dependent upon the accuracy of the State Estimator solution. As experience has accumulated in the use of this tool, it has become evident that in areas of low observability, or low redundancy, RTCA can indicate a need to constrain generation which may not be supported by subsequent analysis.

NERC rules and ERCOT guides do not permit ERCOT to ignore a security/reliability violation; so reducing or eliminating these areas of unobservability has become a priority at ERCOT.

In parallel with our growing experience, ERCOT subcommittees have advanced a proposal to convert the ERCOT market to a Nodal pricing scheme. Several of these proposals are centered upon a successful State Estimator implementation.

In response to one of the proposals at these subcommittees, a white paper, entitled
Telemetry Addendum” was proposed. In this paper, it was stated that:

“ROS will coordinate with WMS to ensure a common understanding of the level of State Estimator performance required to calculate “reasonably accurate LMPs.” The State Estimator performance standard shall define the performance requirements necessary to provide State Estimator results within a desired level of confidence and the calculation of reasonably accurate LMPs. Further, this standard shall address the State Estimator’s ability to detect, correct, and/or otherwise accommodate communications systems failures, failed data points, and missing or inaccurate measurements.”

The existence of this paper was communicated to the Board of Directors on May 18th, 2004.

At its July 13th meeting, in response to this paper, the ROS and ERCOT agreed that ERCOT staff would prepare a draft “Strawman” document describing the technical requirements for data needed for the function of the ERCOT State Estimator. ERCOT suggested that a workshop on State Estimator be held in order to ensure that all the relevant topics were addressed.

ERCOT held two workshops on June 22nd and July 15th to explain its State Estimator methodology and usage and to invite participants to share their related concerns.

This document was prepared in response to the ROS agreement. The document proposes definitions of the required operability and redundancy of telemetry in ERCOT, and explains the proposed levels of observability, redundancy, and analysis to be established for the data requirements for the ERCOT State Estimator.

Executive Summary

The State Estimator is a real-time application used to more accurately and consistently determine the operating state of a large power system. Accurate modeling of the transmission network and sufficient accurate telemetry are vital to this tool.

Maintaining a successful state estimator requires continuous monitoring and management of the validity of thousands of individual measurements. This process is complicated in ERCOT by daily topology changes to the transmission network and accompanying updates to the Network Model at least weekly.

ERCOT is striving to meet the process challenges with an appropriate level of effort; however, ERCOT cannot resolve all issues without coordination with the Transmission Service Providers.

This paper addresses the methodologies ERCOT is to use in determining data required from Transmission Service Providers.

ERCOT proposes a standard that requires observability for all buses which cannot be demonstrated un-necessary to detecting post-contingency overloads. An exception process for demonstrating which bus measurements are un-necessary is proposed and discussed.

ERCOT also proposes and discusses a standard of redundancy which requires redundant measurements at critical locations.

ERCOT staff recommends that this standard be adopted with or without the proposed implementation of nodal pricing in the ERCOT service area.

A summary of these requirements is shown below.

Standard / Exceptions
Observability / All busses 69kV and above unless an exception applies / A bus measurement may be exempted if that measurement can be shown, through network analysis, not needed to calculate any pre- or post-contingency overrating or over limit violation which may result when generation patterns are varied.
Redundancy / Redundancy required for all 345 kV buses and selected 138 kV buses related to critical circuits. Only those 69 kV facilities which are known to, or which may be observed to, have reliability or significant economic impact may be considered for redundancy. Otherwise, 69 kV facilities measurements will be backed up by primitive model values. / -If demonstrated non-critical
-If frequent failure mechanism is corrected.

1. Introduction

1.1 Operating States of a Power System

The operating conditions of a power system at a given point in time may be determined if the network model and complex phasor voltages at every system bus are known. Since a system may be fully defined by the full set of complex phasor voltages, the set of phasor voltages for all buses in a transmission network is referred to as the state of the system. A transmission system may be operating in one of four possible states; namely, secure, insecure, emergency and restorative as operating conditions change.

A power system is said to operate in a secure state if all the loads in the system can be supplied by the on-line generators without violating any operational constraints (i.e. facility ratings and System Operating Limits (SOL)) and if the system can remain in a normal state following the occurrence of any defined contingency.

A state for a transmission interconnection is said to be insecure if the power balance at each bus and all operating inequality constraints are still satisfied, yet the system remains vulnerable with respect to one or more of the considered contingencies. If the system is found to be in a within normal SOLs, but in an insecure operating state, then preventive actions including, but not limited to, re-dispatch of resources, operational switching actions, or other similar actions should be taken to re-position the system and avoid exposure to a move into an emergency state.

Operating conditions may change significantly to cause violation of SOLs. In such a situation, the system is said to be operating in an emergency state. Operating in an emergency state requires immediate corrective action to bring the system back to a secure state without incurring equipment damage. Until such actions are implemented, the system is at increased risk of loss of reliability.

Figure 1.1 – Operating States of a Power System

While a transmission system is in emergency state, corrective control measures must be taken to avoid equipment damage or system collapse. These steps incur expense of re-dispatching resources or disconnecting various loads, lines, transformers or other equipment. As a result, the SOL violations may be eliminated and the transmission systems recover stability with reconfigured topology, re-dispatched resources, and/or reduced load. Such a state is called a restorative state. The actions taken may be referred to as re-posturing of the transmission system

1.2 Why State Estimator?

Although Energy Management Systems (EMS) technology has been used to a certain extent for years, utilities were not pressed to utilize tools that demanded highly accurate real-time network models. Instead, operating margins were established to ensure reliability due to the known lack of accuracy of many models and, thus, study results. This has changed in the energy market environment. In the new environment, the pattern of flows in the network is less predictable than it is in vertically integrated systems, due to new possibilities associated with open access and the operation of the transmission grid under energy market rules.

Although reliability remains the central issue, the need for accurate real-time network models is also crucial for the proper functioning of energy market related functions. These models are based on the results of the State Estimator.

Based on these network models, operators can detect insecure states related to the power grid reliability and respond with corrective action appropriately.

A functioning State Estimator in a control room is analogous to the headlights of an automobile driving on a dark night, it allows the operator to see hazards ahead and avoid them.

What is a State Estimator?

Network real-time models are built from a combination of snapshots of real-time measurements and static network data. Real time measurements consist of analog values and statuses of switching devices. Static network data correspond to representative parameters and basic substation configurations. The real-time model is a mathematical representation of the current conditions in a power network as derived from the State Estimator results.

The State Estimator is used to build the real-time network model. Where an adequate redundancy level exists, the State Estimator will eliminate the effect of bad data (erroneous analog measurements) and allow the temporary loss of measurements without significantly affecting the quality of the estimated values. State Estimation is mainly used to filter redundant data, to eliminate incorrect measurements, and to produce reliable state estimates and, to a certain extent, it allows the determination of power flows in parts of the network that are not directly metered. The quality of results obtained from other EMS network and Market Applications depend on the quality of the real-time network model built via state estimation.

In state estimation, network real-time modeling is decomposed into 1) the processing of statuses of switching devices, and 2) the processing of analog data (power flow, power injections voltage magnitude measurements). During topology processing, the status of breakers/switches are processed using a bus-section/switching device network model. During observability analysis and state estimation, the network topology and parameters are treated as given, and analog data are processed using the bus/branch network model. An incorrect network model would be a major source of problems for state estimation process.

Many portions of the area for which a control center is responsible are normally observable. State estimation may be extended to the rest of the network through the addition of pseudo measurements. Pseudo measurements are calculated values presented to the State Estimator as “measurements”, based on load prediction, using the system total demand and load distribution factors. Use of inaccurate load distribution factors may be another major source of inaccuracies in the state estimation process.

At ERCOT, state estimation is performed every five minute to support both reliability and market applications needs.

Factors Affecting State Estimator Solution Availability and Accuracy

For reasons stated above, the State Estimator must be highly available and provide a solution accurate enough to support the purposes for which it is intended. These do not always happen for several reasons including the following:

a.  Incorrect Network Topology

In state estimation, the network toplogy is treated as given and correct. In the event that the topology is incorrect, the state estimator may not converge, or yield grossly incorrect results. The error in topology may stem from 1) inaccurate statuses of breakers and switching device or 2) error in the primitive network model. Inaccuracies in the statuses of switching devices may be caused by temporary or permanent loss of telemetry.

b.  Inadequate observability

State estimation is extended to the unobservable parts of the network through the addition of pseudo measurements. The pseudo measurements are computed based on load prediction, using load distribution factors. The accuracy of these distribution factors may be questionable when not updated regularly to reflect current conditions. In performing state estimation for the unobservable part of the network, it is possible to corrupt the states estimated from telemetry data

c.  Inadequate redundancy

Redundant measurements are crucial for the detection and identification of bad data. Higher redundancy ensures better reliability of the state estimator solution in the face of temporary loss of measurements.

d.  Incorrect Load distribution factors for unobservable parts of the Network

State Estimator Solution Quality Metrics

The assessment of the current state of the network and the accuracy of security analysis and market functions depends highly on the quality of the state estimator solution. The quality of a SE solution is typically measured using one or more of the following indices

·  Difference between estimated and telemetry values.(i.e weighted residual)

·  Bus MW/MVAR mismatches after SE convergence.

·  Convergence rate of SE.

·  Measurement visibility of network dynamics

The values of these indices depend upon many factors including:

·  Modeling of the electrical devices, connectivity and mapping of telemetries.

·  Availability of telemetry and their quality.

·  Measurement observability of the network.

·  Measurement redundancy of the network. [This term is defined as the ratio of the number of measurements to number of state variables in the observable area of the network.]

·  Communication redundancy.

For more detailed information, State Estimator basics, new technologies and maintenance activities are explained in greater depth in Appendix 1

1.3 Why RTCA? (Real Time Contingency Analysis)

Real Time Contingency Analysis (RTCA) as a core function of system security assessment is critical for detecting problems in a large power system and for determining the current security level of the network. A typical RTCA models single element outages (one transmission line or one transformer), multiple-element outages (two lines or 1 generator and one line) and sequential outages (One outage after another). Checking for over limit results is done after each contingency calculation to determine if the transmission system is secure.